Cheng Cheng

An LLC converter with non-equal-turns-ratios split-transformer (NET-ST) has been proposed to address the large rectified current pulsation in the conventional LLC converters. However, due to the complex resonant processes, the conventional fundamental harmonic approximation is not applicable for analyzing this converter and there is a lack of guidance on designing the NET-ST LLC converter. A data fitting analysis method based on simulation data is proposed to investigate the influence of the parameter K_N (defined as the ratio of two transformers turns ratios, i.e., K_N = n₁/n₂) on characterizing the NET-ST LLC converter. This method establishes the quantitative relationship between the parameter K_N and both the current pulsation and total losses, and reveals the underlying design principles along with the optimal range for K_N. The analysis and experimental results of a 1600W, 400V-input/400V-output prototype demonstrate that the NET-ST design reduces the current pulsation by approximately 40% and losses by about 10% compared to the conventional design. The proposed methodology for analyzing and designing the NET-ST LLC converters has also been promoted to design a split-transformer (ST) LLC named NEI-ST with equal-turns-ratios but non-equal magnetizing inductance. The demonstrated designs validate the effectiveness of presented method and show the appliable of the ST LLC converters.

Fault-tolerant control (FTC) can ensure the three-level T-type inverter (3LT²I) maintains reliable operation when switch is faulty. However, existing FTC strategies exhibit poor output performance and limited neutral-point voltage control (NPVC) ability within the full-range of modulation index (MI) and power factor (PF). To this end, this paper proposes an FTC and NPVC strategy for 3LT²I within full MI and PF range based on modulation signal clamping and decomposition, which improves the system reliability and output performance simultaneously. First, a fundamental cycle is divided into six regions according to the operating conditions when switch fault occurs for independent modulation. Second, the modulation signal is clamped to realize FTC in partial regions and increase NPVC ability in other regions. Furthermore, to realize FTC over the full regions and maximize NPVC ability, the modulation signal is decomposed by selecting the optimal decomposed phase and calculating an accurate decomposition value. The proposed strategy is simple to implement and superior to existing strategies in NPVC ability, output quality and power losses under the full-range of MI and PF. Finally, the effectiveness and superiority of the proposed strategy are validated by experimental results.

The strict absolute stability analysis of the overall grid-following inverter (GFLI) system can be achieved by adopting the orthogonal-normalization-based grid synchronizer (ONGS), but the system robustness is limited by the weak grid, injected power and port voltage level. To improve the absolute stability, this paper proposes a novel grid impedance detection and current adaptive compensation-based ONGS (IDCAC-ONGS), which introduces the grid current feedback signal into the synchronizer input. Based on the absolute stability criteria of the GFLI system composed of synchronous nonlinear criterion and traditional impedance criterion, the reason why the proposed current-compensation-based ONGS (CC-ONGS) has better robustness than the conventional ONGS is explained. Compared with CC-ONGS, a harmonic-injection-based grid impedance detection unit is embedded in proposed IDCAC-ONGS, which can select proper compensation coefficient to further improve the absolute stability. Finally, theoretical analysis and experimental results verify the effectiveness of proposed control strategies.

The Andronov–Hopf virtual oscillator control is a grid-forming control strategy based on instantaneous current feedback, which provides voltage and frequency support in weak grids. However, the unavoidable power coupling and inherent inaccuracies in power tracking control lead to steady-state errors and dynamic oscillations, which severely restrict the power delivery and the grid support capability. Therefore, this article proposes an optimized loop-shaping-based power control (OLPC) strategy to ensure system stability. First, to address the inability of the relative gain array to evaluate feedback dynamics, a closed-loop transfer function array is developed. It explicitly reveals two limitations: failure to achieve power decoupling and amplification of reactive power tracking errors. Subsequently, the objectives of OLPC are formulated in the frequency domain using bandwidth design specifications. Then, the optimal parameters are obtained by reformulating the parameter design problem as a convex optimization problem. Compared to the virtual inductor methods, the OLPC exhibits superior wide-frequency power decoupling, more accurate reactive power tracking, and faster power response. Finally, the theoretical analysis and proposed method are validated through experiments.

Grid-forming inverters (GFMIs) are increasingly being integrated into grid-connected systems dominated by grid-following inverters (GFLIs) to enhance the stability of weak grids. However, the capability of traditional grid-forming control to enhance system stability remains limited. Specifically, when the output power of GFLI increases, the hybrid system consisting of GFMI and GFLI remains susceptible to oscillations and potential instability. To address these limitations, the interaction stability of the hybrid system is comprehensive analyzed and a passivity-enhanced oscillation suppression method is proposed. First, a decomposition impedance model is established to analyze the interaction mechanisms between GFLI and GFMI, as well as the impact of varying power levels. The analysis reveals that increasing the output power of GFMI does not inherently improve its ability to suppress system oscillations. Subsequently, a coupling term responsible for oscillations in the hybrid system is identified and a novel power-adaptive active damping is introduced into the GFMI control to enhance system passivity. In this case, the oscillation of hybrid system can be effectively suppressed under varying power levels and grid impedances. Finally, the experimental results verify the effectiveness of the proposed method.

For high-quality grid connections in distributed generation, the robustness and transient performance of grid-tied inverters (GTI) under various operating conditions are crucial. This article presents a novel model-free adaptive synergetic control (MFASC) approach, which can significantly improve the robustness and transient performance of GTI. To enhance the robustness and adaptability of GTI, and to avoid the effects of unmodeled dynamics, a compact form dynamic linearization model based on the system input/output information is established. Subsequently, in order to improve the response speed, a synergetic control law is designed by utilizing the rapid convergence of differential equations. To further enhance the dynamic performance, an overshoot suppression strategy is designed to suppress the dynamic overshoot through ensuring Lipschitz condition, and a T-factor adaptive law is designed to adjust the response speed and system oscillation by dynamically changing the T parameter of the MFASC. Finally, the MFASC method is applied to the current control of GTI, and the validity is verified through simulation and experiment.

The hybrid modular dc transformer (HMDCT), combining the series resonant dual active bridge (SR-DAB) and phase-shift dual active bridge (PS-DAB), is very suitable for dc bus interlinking applications with bidirectional controllability and high efficiency requirements. However, the coupling between SR-DAB and PS-DAB sub-modules results in a strongly nonlinear dynamic behavior, which in turn poses challenges in instantaneously and coordinately regulating each sub-module to achieve a fast dynamic response of the HMDCT total output. To solve this problem, this article proposes the HMDCT nonlinear input–output large-signal model and a state feedback linearization-based load-current feedforward control (LCFC) strategy. The proposed strategy constructs a linear, load-independent, and easily controllable input–output relationship, while achieving accurate observation of physically nonexistent feedback states in the nonlinear system without relying on complex sampling algorithms and circuits. The system stability including controlled states and uncontrolled internal dynamics is proved, providing a theoretical basis for practical applications. Furthermore, to reduce implementation complexity, the proposed strategy is simplified into improved LCFC form with easily understood physical meaning and less computational effort. Experimental results on a HMDCT prototype show that the proposed strategy significantly improves the dynamic performance in bidirectional operating directions, compared to the conventional feedforward-based strategy.